In recent years, there has been an enormous increase in the amount of digital media, such as music, available online. Services such as Apple's iTunes enable users to legally purchase and download music. Other services such as Yahoo! Music Unlimited and RealNetwork's Rhapsody provide access to millions of songs for a monthly subscription fee. As a result, music has become much more accessible to listeners worldwide. However, the increased accessibility of music has only heightened a long-standing problem for the music industry, which is namely the issue of linking audiophiles with new music that matches their listening preferences.
Many companies, technologies, and approaches have emerged to address this issue of music recommendation. Some companies have taken an analytical approach. They review various attributes of a song, such as melody, harmony, lyrics, orchestration, vocal character, and the like, and assign a rating to each attribute. The ratings for each attribute are then assembled to create a holistic classification for the song that is then used by a recommendation engine. The recommendation engine typically requires that the user first identify a song that he or she likes. The recommendation engine then suggests other songs with similar attributions. Companies using this type of approach include Pandora, SoundFlavor, MusicIP, and MongoMusic (purchased by Microsoft in 2000).
Other companies take a communal approach. They make recommendations based on the collective wisdom of a group of users with similar musical tastes. These solutions first profile the listening habits of a particular user and then search similar profiles of other users to determine recommendations. Profiles are generally created in a variety of ways such as looking at a user's complete collection, the playcounts of their songs, their favorite playlists, and the like. Companies using this technology include Last.fm, Music Strands, WebJay, Mercora, betterPropaganda, Loomia, eMusic, musicmatch, genielab, upto11, Napster, and iTunes with its celebrity playlists.
The problem with these traditional recommendation systems is that they fail to consider peer influences. For example, the music that a particular teenager listens to may be highly influenced by the music listened to by a group of the teenager's peers, such as his or her friends. As such, there is a need for a music recommendation system and method that recommends music to a user based on the listening habits of a peer group.